In order to better carry out the monitoring and early warning of hail weather in Jiangxi, MICAPS weather map, Jiangxi meteorological automatic station data, Jiangxi radar mosaic data, single radar PUP product data, double multi wind field data inversion, strong echo automatic recognition, hail disaster photos and WeChat feedback video and other data are used, and the principles and methods of synoptic, radar meteorology, image recognition technology, etc. are adopted, The echo characteristics of hail in Jiangxi and surrounding provinces on March 14, 2022 are analyzed. The results show that: ① On March 14, hail occurred in many places in Jiangxi, strong wind ≥ 17.2 m/s appeared in 20 regional stations, precipitation ≥ 50.0 mm appeared in 24 regional stations, and thunderstorm strong wind and cold air strong wind mixed. ② 200 hPa high-altitude outflow area, 500 hPa south branch and wind shear, 850 hPa low vortex and shear line, ground convergence line and southwest inverted trough are obvious features of weather background; The higher CAPE value in Nanchang, the stronger inversion layer, the middle layer dry area and the lower layer wet area provide the environmental conditions for hail weather. ③ Hail is dominated by supercell echo, sometimes isolated, sometimes in echo group and echo band; Echo intensity ≥ 60 dBz and ≥ 70 dBz; Strong echo area: minimum ≥ 18 km2, maximum ≥ 180 km2; 30-60 dBz range strong echo gradient ≤ 6 km, with obvious cloud anvil echo structure; The life history of hail echo is more than 2 hours. ④ In hail echo recognition, vertical integration of liquid water content VIL is a very important feature. The VIL of Jiangxi hail single radar is 35 ~ 65 kg/m2, and the VIL on the radar puzzle is ≥ 35 kg/m2. ⑤ On the CAPPI chart of 2.5 km hail echo, the hail echo intensity is ≥ 60 dBz, and the maximum is 65 dBz; Most echo centers on double multi wind fields have the characteristics of middle vortex structure, side convergence wind field, north south wind field convergence, etc; Some wind farms are messy. ⑥ Scatter contour algorithm is adopted for automatic echo edge recognition; In each recognition, the thunderstorm echo group can identify multiple strong echoes at the same time, up to 19; ≥ 60 dBz echo area: minimum 18 km2, maximum 184 km2; Most cases showed echo nuclei of 65 dBz, the strongest being 73 dBz; The minimum and maximum echo gradients of 16 hailstorms are 1 km and 5.66 km respectively; Most cloud anvil echoes are between 1:1 and 3. The above analysis results provide an analysis basis for the monitoring and early warning of hail weather in Jiangxi.